The objective of this thesis is to evaluate different models and methods for website classification. The websites are classified based on their functionality, in this case specifically whether they are forums, news sites or blogs. The analysis aims at solving a search engine problem, which means that it is interesting to know from which categories in a information search the results come. The data consists of two datasets, extracted from the web in January and April 2013. Together these data sets consist of approximately 40.000 observations, with each observation being the extracted text from the website. Approximately 7.000 new word variables were subsequently created from this text, as were variables based on Latent Dirichlet Allocation. One variable (the number of links) was created using the HTML-code for the web site. These data sets are used both in multinomial logistic regression with Lasso regularization, and to create a Naive Bayes classifier. The best classifier for the data material studied was achieved when using Lasso for all variables with multinomial logistic regression to reduce the number of variables. The  accuracy of this model is 99.70 %. When time dependency of the models is considered, using the first data to make the model and the second data for testing, the accuracy, however, is only 90.74 %. This indicates that the data is time dependent and that websites topics change over time.

The duration of children stay in a temporary foster care system needs to be minimal in order to ensure a stable and successful life. However, a time-consuming procedure of investigations is usually taken to decide whether they can reunite with their birth parents. Moreover, if the child fails to reunite with their family, another discharge decision needs to be assessed, leading to even longer time without a normal life. Based on the data from Adoption and Foster Care Analysis and Reporting System (AFCARS), this thesis derives a prediction model to discriminate the children with a tendency of unsuccessful reunification from the rest. An alternative discharge option can therefore be prepared concurrently for the foster youth with high non-reunification probability. The model is obtained by logistic regression and evaluated with receiver operating characteristic (ROC) curve.

The association between the risk of liver cancer and blood chemistry was investigated in a cohort study with 95,150 men and women from two counties in Sweden. In 1963-65, blood tests and physical measurements were undertaken. All individuals were then followed up until 2007, and a total of 312 were diagnosed with liver cancer. Using survival analysis and logistic regression, significant risk factors were identified. Stepwise Cox proportional hazards regression applied to a main effect model revealed that Glutamic Pyruvate Transaminase (GPT) and Thymol Turbidity (TYM) were the most significant risk factors (p<0.0001), followed by Protein-Bound Hexoses (HEX)  (p=0.002), sex (p=0.02), and Serum Iron (p= 0.03). Increasing the level of GPT expressed in U/L from normal (<21) to slightly elevated (21, 31) or substantially elevated (>31) raised the hazard of experiencing liver cancer by a factor of 1.45 and 4.09, respectively. In addition, GPT was found to be the most significant risk factor in almost all age groups among both men and women. However, there was no evidence that elevated GPT levels within the normal range (<21), influenced the risk of liver cancer. Additional subgroup analyses revealed that TYM was highly significant within the group with normal GPT, and a high level of HEX (≥134 mg/dl) increased the hazard 1.55 times in comparison with the lowest HEX group (<115 mg/dl). BMI was significant only in the male subgroup  (p<0.01) and, in the obesity group, the hazard of experiencing liver cancer was 1.99 times higher than in the normal BMI group. A significant three-way interaction between GPT, BMI and gender was present (p=0.05) with a robust significant two-way interaction between GPT and BMI (p<0.01) in the male subgroup.

This study discusses main risk factors of ovarian tumor to determine a tumor type. Since symptoms of ovarian tumor are not obvious as the tumors are located in pelvic, the ovarian tumor is difficult to detect. The symptoms are mostly stomach or lower abdomen swellings, which are often ignored. The probability of ovarian cancer is lower than cervical cancer, but the mortality rate is the highest of all gynecologic diseases. The study uses statistical methods to analyze risk factors of patients to determine the tumor type and an early treatment in order to reduce the death rate. The sources of the studies are from Kaohsiung Veterans General Hospital and are classified according to different cases of tumors based on ultrasound checks and other relevant risk factors, such as ages and tumor marks so as to provide a determined method to distinguish among benign, borderline and malignant ovarian tumors in order to create appropriate classification criteria for followups, surgeries, and references for tracking. To differentiate between malignant and nonmalignant (benign and borderline) cases, we use risk factors to construct classification and regression trees so as to help physicians to determine the tumor type. In the situation in which the non-malignant tumor may be determined, we use logistic regression model according to the degree of influence of risk factors to further classify between benign and borderline tumors. The aforementioned process can determine tumor types precisely and can also determine surgery types so as to help determining whether patients would need a follow-up.

NO

none

none

The research of corporate financial distress prediction model is always one of the important topics in financial management; and mostly people do the research and extract sample companies based on the definition for corporate default by Taiwan Economic Journal. However, we think the timing to observe the potential corporate financial distress is extremely vital; the actual benefit will not be good even with high accuracy if relevant counterparties recognize it too late to undertake certain action for mitigating loss. The main purpose of this study is trying to alert potential corporate financial distress as early as possible, and then could contribute some to this topic. This study extracts 34 financial alerted sample companies with share prices plumped by 50% dramatically or alternatively with share prices diminished below their face value while the stock market index rose in 2007. We matched each sample company by another financially healthy company from the same industry, chose 25 financial ratios to be the variables, and running through each year by adopting logistic regression analysis. We put all variables into the regression formula and weeded out insignificant prediction variables one by one by Wald Backward Elimination, and then sieved out relatively meaningful ones. The first conclusion of this study is that we should use quarter as the financial intervals for this type of sample companies. Secondly, we found that in December and September 2007 there were three significant variables, i.e. Return on Equity (ROE), net income, operational profit ratio, inventory and account receivable to equity ratio. Thirdly, there were three significant variables in June 2007, i.e. earning before tax ratio, growth ratio of operational profit and total liability/ total equity.

This paper provides a credit risk quantification system for banks to estaminate the credit risk of loans to small and mediume nterprises(SMEs). As we know, the most difficult thing for banks to handle SME loans is whose financial reporting lacks transparency and no valuable reference. We use non-financial variables and employ the logisitic regression to develop the credit risk predict model. We concludet: first, when construct a SMEs credit rating system, non-financial factors should be seriously considered and adopted. Second, because of positioned different stage of firm life cycle, the credit rating model should be set up differently by different stage of firm. Third, SME loans should to make much of establishing ¡§relationship-based¡¨ in order to meet the various demands of risk management.

In the late 90s coconut farms in Mozambique were affected by a disease that made the coconut trees drop their leaves and die, the disease is called the coconut lethal yellowing disease (CLYD). It is known that planthoppers are spreading the disease. This thesis investigates if cultivation and farm related factors could have an impact on the risk of being infected by CLYD. With a sample of 534 farms from the two provinces Zambeze and Nampula a logistic regression model is estimated. The result shows that the only factor that has a significant effect of increasing the risk of getting infected by the disease is if farms had other palm species than coconut trees on the plantation. / I slutet av 90-talet drabbades många kokosnötsfarmer i Mocambique av en sjukdom som gjorde att träden började tappa sina löv och till slut dog. Denna sjukdom kallas för ”the coconut lethal yellowing disease” (CLYD). I dagsläget vet man att det är en specifik insekt som sprider denna sjukdom. Denna uppsats undersöker om det finns andra faktorer som kan påverka risken för koksnötsfarmer att drabbas av denna sjukdom, faktorer som är kopplade till farmerna och dess vegetation. Genom ett urval av 534 farmer från de två provinserna Zambezia och Nampula skattades en logistik regression. Resultatet visade att det enbart är en faktor, om plantagen innehåller andra palmsorter, som signifikant ökar risken för att en farm ska drabbas av CLYD.